{smcl}
{com}{sf}{ul off}{txt}{.-}
      name:  {res}<unnamed>
       {txt}log:  {res}C:\Users\jp18390\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Age Discrimination Project\PAR R&R\Statistics\Age_Discrimination.APPENDIX 2.NONLINEAR FUNCTIONAL FORM ANALYSES.06-02-2022.smcl
  {txt}log type:  {res}smcl
 {txt}opened on:  {res} 2 Jun 2022, 18:06:59
{txt}
{com}.    
.    
. 
. use "C:\Users\jp18390\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Age Discrimination Project\PAR R&R\Statistics\Age_Discrimination.Post-Estimation.06-02-2022.dta", replace 
{txt}
{com}. 
. 
. 
. 
. **** 2022 PAR Data Replication [06/02/2022]: "STATUS-GROUP POWER DIFFERENTIALS AND AGE DISCRIMINATION WITHIN THE U.S. FEDERAL WORKFORCE" [KRAUSE & PARK] ****
. 
.  
. 
. 
. 
.  
. **** APPENDIX SECTION 2: COLLINEARITY OF HIGHER POWERS AND LOWER MODEL FIT OF NONLINEAR FUNCTIONAL FORM RELATING TO PRIMARY COVARIATES OF INTEREST *****
.  
.  
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
.  
. xtset a_id year, yearly
{res}{txt}{col 8}panel variable:  {res}a_id (unbalanced)
{txt}{col 9}time variable:  {res}{col 25}year, 2010 to 2019, but with gaps
{txt}{col 17}delta:  {res}1 year
{txt}
{com}. 
. *
. * 
.  
.  
. ***  APPENDIX SECTION A2.1A: BIVARIATE CORRELATIONS AND VIF ANALYSIS FROM BETWEEN-EFFECTS MODELS [BASIS FOR MAIN EVIDENCE CONSISTENT WITH STATUS-GROUP POWER DIFFERENTIALS] ***
. 
. 
. 
. *** STEP 1: COLLAPSE DATASET INTO BETWEEN-EFFECTS [AGENCY GROUP MEANS] FOR PURPOSES OF EVALUATING AGENCY-MEANS OF ALL VARIABLES **
. 
. collapse age_discrimination ratio_40over_suplb  ratio_40over_nonsuplb  ratio40suplb_nonsuplb      orgjustice_sem  nonprof40over_tr_lb  ///
> politicization_lb   ratio_fsup_msup ratio_minsup_nonmsup lntwf, by(a_id)
{txt}
{com}. *
. *
. *
. ** STEP 2: EVALUATE COLLINEARITY ARISING FROM HIGHER POWERS OF 'ORGANIZATIONAL SUSCEPTIBILITY' COVARIATES USING BIVARIATE CORRELATIONS AND VIFs **
. 
. ** STEP 2a: FOR THE PROPORTION OF 'OLDER' SUPERVISORY PERSONNEL WITHIN AN AGENCY **
. gen ratio_40over_suplb_quad  =  ratio_40over_suplb^2
{txt}
{com}. gen ratio_40over_suplb_cubic =  ratio_40over_suplb^3
{txt}
{com}. *
. *
. correlate ratio_40over_suplb ratio_40over_suplb_quad ratio_40over_suplb_cubic 
{txt}(obs=130)

             {c |} r~_suplb ratio_~d ratio_~c
{hline 13}{c +}{hline 27}
ratio~_suplb {c |}{res}   1.0000
{txt}ratio_40ov~d {c |}{res}   0.9982   1.0000
{txt}ratio_40ov~c {c |}{res}   0.9934   0.9985   1.0000

{txt}
{com}. *
. regress age_discrimination ratio_40over_suplb ratio_40over_suplb_quad ratio_40over_suplb_cubic 

{txt}      Source {c |}       SS           df       MS      Number of obs   ={res}       130
{txt}{hline 13}{c +}{hline 34}   F(3, 126)       = {res}     1.25
{txt}       Model {c |} {res} 15621.0345         3  5207.01149   {txt}Prob > F        ={res}    0.2942
{txt}    Residual {c |} {res} 524438.353       126  4162.20915   {txt}R-squared       ={res}    0.0289
{txt}{hline 13}{c +}{hline 34}   Adj R-squared   ={res}    0.0058
{txt}       Total {c |} {res} 540059.388       129  4186.50688   {txt}Root MSE        =   {res} 64.515

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 1}age_discri~n{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
ratio~_suplb {c |}{col 14}{res}{space 2}-1136.976{col 26}{space 2} 14504.22{col 37}{space 1}   -0.08{col 46}{space 3}0.938{col 54}{space 4}-29840.41{col 67}{space 3} 27566.46
{txt}ratio_40ov~d {c |}{col 14}{res}{space 2}  2513.09{col 26}{space 2} 18199.16{col 37}{space 1}    0.14{col 46}{space 3}0.890{col 54}{space 4}-33502.51{col 67}{space 3} 38528.69
{txt}ratio_40ov~c {c |}{col 14}{res}{space 2}-1514.075{col 26}{space 2}  7558.26{col 37}{space 1}   -0.20{col 46}{space 3}0.842{col 54}{space 4}-16471.65{col 67}{space 3}  13443.5
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 115.6885{col 26}{space 2} 3825.334{col 37}{space 1}    0.03{col 46}{space 3}0.976{col 54}{space 4}-7454.535{col 67}{space 3} 7685.912
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. estat vif

{txt}    Variable {c |}       VIF       1/VIF  
{hline 13}{c +}{hline 22}
ratio_40ov~d {c |} {res}115363.49    0.000009
{txt}ratio_40ov~c {c |} {res} 31953.49    0.000031
{txt}ratio~_suplb {c |} {res} 26212.75    0.000038
{txt}{hline 13}{c +}{hline 22}
    Mean VIF {c |} {res} 57843.24
{txt}
{com}. *
. *
. *
. ** STEP 2b: FOR THE PROPORTION OF 'OLDER' NON-SUPERVISORY PERSONNEL WITHIN AN AGENCY **
. gen ratio_40over_nonsuplb_quad  =  ratio_40over_nonsuplb^2
{txt}
{com}. gen ratio_40over_nonsuplb_cubic =  ratio_40over_nonsuplb^3
{txt}
{com}. *
. *
. correlate ratio_40over_nonsuplb ratio_40over_nonsuplb_quad ratio_40over_nonsuplb_cubic 
{txt}(obs=130)

             {c |} ratio_.. ratio_.. ratio_..
{hline 13}{c +}{hline 27}
r~r_nonsuplb {c |}{res}   1.0000
{txt}~nsuplb_quad {c |}{res}   0.9960   1.0000
{txt}ratio_40ov.. {c |}{res}   0.9845   0.9962   1.0000

{txt}
{com}. *
. regress age_discrimination ratio_40over_nonsuplb ratio_40over_nonsuplb_quad ratio_40over_nonsuplb_cubic 

{txt}      Source {c |}       SS           df       MS      Number of obs   ={res}       130
{txt}{hline 13}{c +}{hline 34}   F(3, 126)       = {res}     0.39
{txt}       Model {c |} {res} 4911.56881         3   1637.1896   {txt}Prob > F        ={res}    0.7636
{txt}    Residual {c |} {res} 535147.819       126  4247.20491   {txt}R-squared       ={res}    0.0091
{txt}{hline 13}{c +}{hline 34}   Adj R-squared   ={res}   -0.0145
{txt}       Total {c |} {res} 540059.388       129  4186.50688   {txt}Root MSE        =   {res} 65.171

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 1}age_discri~n{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
r~r_nonsuplb {c |}{col 14}{res}{space 2}-1191.402{col 26}{space 2} 5235.126{col 37}{space 1}   -0.23{col 46}{space 3}0.820{col 54}{space 4}-11551.56{col 67}{space 3} 9168.758
{txt}~nsuplb_quad {c |}{col 14}{res}{space 2}  2165.57{col 26}{space 2} 7828.876{col 37}{space 1}    0.28{col 46}{space 3}0.783{col 54}{space 4}-13327.55{col 67}{space 3} 17658.69
{txt}ratio_40ov.. {c |}{col 14}{res}{space 2} -1278.44{col 26}{space 2} 3857.254{col 37}{space 1}   -0.33{col 46}{space 3}0.741{col 54}{space 4}-8911.831{col 67}{space 3} 6354.952
{txt}{space 7}_cons {c |}{col 14}{res}{space 2} 239.6485{col 26}{space 2} 1152.334{col 37}{space 1}    0.21{col 46}{space 3}0.836{col 54}{space 4}-2040.786{col 67}{space 3} 2520.083
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. estat vif

{txt}    Variable {c |}       VIF       1/VIF  
{hline 13}{c +}{hline 22}
~nsuplb_quad {c |} {res} 26204.69    0.000038
{txt}ratio_40ov.. {c |} {res}  6863.29    0.000146
{txt}r~r_nonsuplb {c |} {res}  6409.80    0.000156
{txt}{hline 13}{c +}{hline 22}
    Mean VIF {c |} {res} 13159.26
{txt}
{com}. *
. *
. *
. *
. ** STEP 2C: FOR THE RATIO OF THE PROPORTION OF 'OLDER' SUPERVISORY PERSONNEL TO 'OLDER' NON-SUPERVISORY PERSONNEL WITHIN AN AGENCY **
. gen ratio40suplb_nonsuplb_quad  =  ratio40suplb_nonsuplb ^2
{txt}
{com}. gen ratio40suplb_nonsuplb_cubic =  ratio40suplb_nonsuplb ^3
{txt}
{com}. 
. *
. *
. correlate ratio40suplb_nonsuplb  ratio40suplb_nonsuplb_quad ratio40suplb_nonsuplb_cubic 
{txt}(obs=130)

             {c |} ratio4~b ratio4~d ratio4~c
{hline 13}{c +}{hline 27}
ratio40sup~b {c |}{res}   1.0000
{txt}ratio40sup~d {c |}{res}   0.9928   1.0000
{txt}ratio40sup~c {c |}{res}   0.9716   0.9928   1.0000

{txt}
{com}. *
. regress age_discrimination ratio40suplb_nonsuplb  ratio40suplb_nonsuplb_quad ratio40suplb_nonsuplb_cubic 

{txt}      Source {c |}       SS           df       MS      Number of obs   ={res}       130
{txt}{hline 13}{c +}{hline 34}   F(3, 126)       = {res}     0.38
{txt}       Model {c |} {res} 4873.31485         3  1624.43828   {txt}Prob > F        ={res}    0.7658
{txt}    Residual {c |} {res} 535186.073       126  4247.50851   {txt}R-squared       ={res}    0.0090
{txt}{hline 13}{c +}{hline 34}   Adj R-squared   ={res}   -0.0146
{txt}       Total {c |} {res} 540059.388       129  4186.50688   {txt}Root MSE        =   {res} 65.173

{txt}{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 1}age_discri~n{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
ratio40sup~b {c |}{col 14}{res}{space 2} 1143.541{col 26}{space 2} 1552.516{col 37}{space 1}    0.74{col 46}{space 3}0.463{col 54}{space 4}-1928.842{col 67}{space 3} 4215.925
{txt}ratio40sup~d {c |}{col 14}{res}{space 2}-763.8427{col 26}{space 2}  1133.23{col 37}{space 1}   -0.67{col 46}{space 3}0.502{col 54}{space 4}-3006.472{col 67}{space 3} 1478.787
{txt}ratio40sup~c {c |}{col 14}{res}{space 2} 161.5332{col 26}{space 2}  271.865{col 37}{space 1}    0.59{col 46}{space 3}0.553{col 54}{space 4}-376.4796{col 67}{space 3} 699.5461
{txt}{space 7}_cons {c |}{col 14}{res}{space 2}-522.6287{col 26}{space 2} 700.7433{col 37}{space 1}   -0.75{col 46}{space 3}0.457{col 54}{space 4}-1909.379{col 67}{space 3} 864.1216
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. estat vif

{txt}    Variable {c |}       VIF       1/VIF  
{hline 13}{c +}{hline 22}
ratio40sup~d {c |} {res}  6870.68    0.000146
{txt}ratio40sup~b {c |} {res}  1763.24    0.000567
{txt}ratio40sup~c {c |} {res}  1756.80    0.000569
{txt}{hline 13}{c +}{hline 22}
    Mean VIF {c |} {res}  3463.58
{txt}
{com}. 
. 
. 
. save "C:\Users\jp18390\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Age Discrimination Project\PAR R&R\Statistics\collapsedagediscrim1.06-02-2022.dta", replace 
{txt}file C:\Users\jp18390\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Age Discrimination Project\PAR R&R\Statistics\collapsedagediscrim1.06-02-2022.dta saved

{com}. *
. *
. *
. *
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
.  
. 
.  
. ***  APPENDIX SECTION A2.1B: EVALUATE LINEAR VERSUS SQUARED POWERS BASED ON MODEL FIT (BIC) STATISTICS FROM HYBRID PANEL MODELS ***
. 
. 
. 
. use "C:\Users\jp18390\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Age Discrimination Project\PAR R&R\Statistics\Age_Discrimination_Dataset_06-02-2022.dta", replace 
{txt}
{com}. 
. 
. xtset a_id year, yearly
{res}{txt}{col 8}panel variable:  {res}a_id (unbalanced)
{txt}{col 9}time variable:  {res}{col 25}year, 2010 to 2019, but with gaps
{txt}{col 17}delta:  {res}1 year
{txt}
{com}. 
. *
. *
. 
. ** STEP 1: CREATE VALUATE LINEAR VERSUS SQUARED POWERS POWERS BASED ON MODEL FIT (BIC) STATISTICS:  **
. 
. gen ratio_40over_suplb_quad  =  ratio_40over_suplb^2
{txt}
{com}. *   
. gen ratio_40over_nonsuplb_quad  =  ratio_40over_nonsuplb^2
{txt}
{com}. *
. gen ratio40suplb_nonsuplb_quad  =  ratio40suplb_nonsuplb ^2
{txt}
{com}.  
. 
. 
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
.  
. 
. 
. ** MODEL A2.1: DISAGGREGATE SUPERVISOR/SUBORDINATE RATIO MEASURES [ratio_40over_suplb; ratio_40over_nonsuplb]: ONLY RANDOM INTERCEPT MODEL SPECIFICATION WITH BE & WE ESTIMATES FOR ALL COVARIATES [SANS YEAR UNIT EFFECTS & LN(TOTAL WORKFORCE)] --- REDUCED MODEL: ONLY CONTROL COVARIATES ARE LNTEF & YEAR UNIT EFFECTS ***
. 
. *
. *** LINEAR FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio_40over_suplb  ratio_40over_nonsuplb   lntwf ///
> yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10 ratio_55over_totallb , clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio_40over_suplb  ratio_40over_nonsuplb  ratio_55over_totallb) 
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2761.055{col 49}    19{col 58}  5560.11{col 69} 5651.292
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. 
. *
. *
. *
. *** QUADRATIC FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio_40over_suplb  ratio_40over_suplb_quad   ratio_40over_nonsuplb  ratio_40over_nonsuplb_quad  lntwf ///
> yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10  ratio_55over_totallb , clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio_40over_suplb ratio_40over_suplb_quad   ratio_40over_nonsuplb  ratio_40over_nonsuplb_quad  ratio_55over_totallb) 
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2759.078{col 49}    23{col 58} 5564.157{col 69} 5674.535
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. *
. *
. *
. 
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. ****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. 
. 
. 
. ** MODEL A2.2: DISAGGREGATE SUPERVISOR/SUBORDINATE RATIO MEASURES [ratio_40over_suplb; ratio_40over_nonsuplb]: ONLY RANDOM INTERCEPT MODEL SPECIFICATION WITH BE & WE ESTIMATES FOR ALL COVARIATES [SANS YEAR UNIT EFFECTS & LN(TOTAL WORKFORCE)] --- FULL MODEL: ALL CONTROL COVARIATES ARE LNTEF & YEAR UNIT EFFECTS ***
. 
. *** LINEAR FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio_40over_suplb  ratio_40over_nonsuplb   orgjustice_sem   nonprof40over_tr_lb   politicization_lb ratio_fsup_msup ratio_minsup_nonmsup lntwf ///
> yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10 ratio_55over_totallb , clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio_40over_suplb  ratio_40over_nonsuplb    orgjustice_sem  nonprof40over_tr_lb   politicization_lb   ratio_fsup_msup ratio_minsup_nonmsup ratio_55over_totallb) 
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38} -2738.61{col 49}    29{col 58} 5535.221{col 69} 5674.393
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. *
. *
. *
. *
. *
. *** QUADRATIC FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio_40over_suplb  ratio_40over_suplb_quad   ratio_40over_nonsuplb  ratio_40over_nonsuplb_quad  orgjustice_sem   nonprof40over_tr_lb politicization_lb ratio_fsup_msup ratio_minsup_nonmsup lntwf yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10  ratio_55over_totallb, clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full  use(ratio_40over_suplb  ratio_40over_suplb_quad   ratio_40over_nonsuplb  ratio_40over_nonsuplb_quad  orgjustice_sem  nonprof40over_tr_lb   politicization_lb ratio_fsup_msup ratio_minsup_nonmsup ratio_55over_totallb) 
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2737.682{col 49}    33{col 58} 5541.364{col 69} 5699.733
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. *
. *
. *
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. ****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. 
. 
. 
. ** MODEL A2.3: RATIO OF OVER40 SUPERVISORS TO OVER40 NON-SUPERVISORS [RATIO OF ratio_40over_suplb TO ratio_40over_nonsuplb]: ONLY RANDOM INTERCEPT MODEL SPECIFICATION WITH BE & WE ESTIMATES FOR ALL COVARIATES [SANS YEAR UNIT EFFECTS & LN(TOTAL WORKFORCE)] REDUCED MODEL: ONLY CONTROL COVARIATES ARE LNTEF & YEAR UNIT EFFECTS ***
. 
. *** LINEAR FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio40suplb_nonsuplb   lntwf ///
> yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10  ratio_55over_totallb, clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio40suplb_nonsuplb ratio_55over_totallb) 
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2761.816{col 49}    17{col 58} 5557.632{col 69} 5639.216
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. *
. *
. *
. 
. *** QUADRATIC FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio40suplb_nonsuplb  ratio40suplb_nonsuplb_quad  lntwf ///
> yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10  ratio_55over_totallb, clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio40suplb_nonsuplb  ratio40suplb_nonsuplb_quad  ratio_55over_totallb) 
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2761.018{col 49}    19{col 58} 5560.035{col 69} 5651.217
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. *
. *
. *
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. ****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. 
. 
. 
. ** MODEL A2.4: RATIO OF OVER40 SUPERVISORS TO OVER40 NON-SUPERVISORS [RATIO OF ratio_40over_suplb TO ratio_40over_nonsuplb]: ONLY RANDOM INTERCEPT MODEL SPECIFICATION WITH BE & WE ESTIMATES FOR ALL COVARIATES [SANS YEAR UNIT EFFECTS & LN(TOTAL WORKFORCE)] --- FULL MODEL: ALL CONTROL COVARIATES ARE LNTEF & YEAR UNIT EFFECTS ***
. 
. *** LINEAR FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio40suplb_nonsuplb    orgjustice_sem    nonprof40over_tr_lb     politicization_lb ratio_fsup_msup ratio_minsup_nonmsup lntwf ///
> yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10 ratio_55over_totallb , clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio40suplb_nonsuplb   orgjustice_sem   nonprof40over_tr_lb  politicization_lb   ratio_fsup_msup ratio_minsup_nonmsup ratio_55over_totallb)  
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2738.135{col 49}    27{col 58}  5530.27{col 69} 5659.844
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. 
. *
. *
. *
. *
. 
. *** QUADRATIC FUNCTIONAL FORM ***
. quietly xthybrid age_discrimination  ratio40suplb_nonsuplb   ratio40suplb_nonsuplb_quad   orgjustice_sem    nonprof40over_tr_lb     politicization_lb ratio_fsup_msup  /// 
> ratio_minsup_nonmsup lntwf yr2 yr3 yr4 yr5 yr6 yr7 yr8 yr9 yr10  ratio_55over_totallb, clusterid(a_id) vce(cluster a_id)  family(nbinomial) link(log) full ///
> use(ratio40suplb_nonsuplb  ratio40suplb_nonsuplb_quad    orgjustice_sem   nonprof40over_tr_lb  politicization_lb   ratio_fsup_msup ratio_minsup_nonmsup ratio_55over_totallb)  
{txt}
{com}. *
. estat ic

Akaike's information criterion and Bayesian information criterion

{txt}{hline 13}{c TT}{hline 63}
       Model {c |}        Obs  ll(null)  ll(model)      df         AIC        BIC
{hline 13}{c +}{hline 63}
{ralign 12:{stata estimates replay model:model}}{col 14}{c |}{res}{col 16}       897{col 27}        .{col 38}-2737.213{col 49}    29{col 58} 5532.425{col 69} 5671.598
{txt}{hline 13}{c BT}{hline 63}
{p 15 21 2}
Note: N=Obs used in calculating BIC; see {helpb bic_note:[R] BIC note}.
{p_end}

{com}. *
. *
. *
. 
. 
. 
. 
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. ****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. *****************************************************************************************************************************************************************************************
. 
. log close
      {txt}name:  {res}<unnamed>
       {txt}log:  {res}C:\Users\jp18390\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Age Discrimination Project\PAR R&R\Statistics\Age_Discrimination.APPENDIX 2.NONLINEAR FUNCTIONAL FORM ANALYSES.06-02-2022.smcl
  {txt}log type:  {res}smcl
 {txt}closed on:  {res} 2 Jun 2022, 18:07:10
{txt}{.-}
{smcl}
{txt}{sf}{ul off}